[ Back to EurekAlert! ]

PUBLIC RELEASE DATE:
29-Jul-2014

[ | E-mail ] Share Share

Contact: Iqbal Pittalwala
iqbal@ucr.edu
951-827-6050
University of California - Riverside
www.twitter.com/UCRiverside

1996 research article deemed a classic paper

Paper by UC Riverside's Michael Pazzani recognized for personalizing Internet content and learning user profiles

IMAGE: Michael Pazzani is the vice chancellor for research and economic development at UC Riverside.

Click here for more information.

RIVERSIDE, Calif. — A 1996 research paper authored by University of California, Riverside's Michael J. Pazzani and two colleagues has been selected by the Association for the Advancement of Artificial Intelligence (AAAI) to win the 2014 Classic Paper Award.

The AAAI, which promotes theoretical and applied artificial intelligence research, established the award in 1999 to honor authors of papers, chosen from a specific conference year, that were deemed most influential.

"The Classic Paper Award this year is given to the paper deemed most influential from The Thirteenth National Conference on Artificial Intelligence held in Portland, Oregon in 1996," said Manuela Velosa, the president of the AAAI. "Pazzani and his colleagues are being recognized for significant contributions to the field of personalizing Internet content and learning user profiles."

Pazzani, the first author of the research paper, wrote it when he was a professor of information & computer science and cognitive science at UC Irvine. Now the vice chancellor for research and economic development at UC Riverside, he received the award today (July 29) at The Twenty-eighth AAAI Conference on Artificial Intelligence (AAAI-14) in Québec City, Canada, during the opening ceremony.

The research paper in question is "Syskill & Webert: Identifying Interesting Web Sites," published in the proceedings of The Thirteenth National Conference on Artificial Intelligence (AAAI-96). Pazzani's coauthors are Jack Muramatsu and Daniel Billsus.

Their paper showed how a profile of the user can be learned by any learning algorithm from a user's feedback on any web page and how this profile can be used to predict the user's interest in web pages. By combining the system with a search engine, the paper showed how search results can be personalized and how a query can be constructed to search for content that interests the user.

"The mid-1990s were an exciting time to be a computer scientist studying machine learning," Pazzani said. "The World Wide Web was a promising area for research and just starting to take off. My colleagues and I developed one of the first general purpose systems for learning about a user's interests from the user's web browsing behavior."

Pazzani explained that today many websites personalize the presentation of content to individual users. For example, sites such as Yahoo News learn user interests. YouTube has thumbs up and thumbs down buttons for users to rate content, and uses this feedback to recommend videos.

"In 1996, however, the internet experience was a one-size fits all experience," he said. "Now, the personalization of content may be the most common application of artificial intelligence encountered by the average person."

The research paper directly led to commercial applications. Pazzani and Billsus continued to develop and refine content personalization algorithms in the next few years and founded a company that personalized content for several content providers utilizing the same basic approach for users. The commercial application of the personalization was deployed on the Los Angeles Times mobile news and classified ads sites. (Later Pazzani and Billsus sold the company.)

"The project is an ideal illustration of the value of research at University of California," Pazzani said.

The research paper was the first in the AAAI conferences to explore content-based personalization during internet browsing and searching. Having withstood the test of time, it is among the most cited from AAAI-96. Noteworthy for its experimental approach to evaluating several learning algorithms, it has been cited especially by researchers exploring more advanced approaches for collecting feedback, representing documents, and learning profiles.

###

The University of California, Riverside is a doctoral research university, a living laboratory for groundbreaking exploration of issues critical to Inland Southern California, the state and communities around the world. Reflecting California's diverse culture, UCR's enrollment has exceeded 21,000 students. The campus opened a medical school in 2013 and has reached the heart of the Coachella Valley by way of the UCR Palm Desert Center. The campus has an annual statewide economic impact of more than $1 billion. A broadcast studio with fiber cable to the AT&T Hollywood hub is available for live or taped interviews. UCR also has ISDN for radio interviews. To learn more, call (951) UCR-NEWS.



[ Back to EurekAlert! ] [ | E-mail Share Share ]

 


AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert! system.